What to do if you need to serialize data that is not supported by the standard json module?

You can use the following approaches:

• Extend the standard JSONEncoder serializer. This allows you to define how to serialize custom data types.

  import json
  
  class CustomEncoder(json.JSONEncoder):
      def default(self, o):
          if isinstance(o, set):
              return list(o)  # Serialize sets as lists
          if hasattr(o, '__dict__'):
              return o.__dict__ # Serialize objects with attributes as dictionaries
          return super().default(o)
  

• Use external libraries. There are libraries that provide more flexible and powerful serialization tools.

  Library	Description	Example usage
  pickle	Serializes and deserializes Python objects. Unsafe for external data.	import pickle; data = {'a': 1}; pickled_data = pickle.dumps(data)
  jsonpickle	Extends the standard json, supporting more complex types.	import jsonpickle; data = {1, 2}; json_string = jsonpickle.encode(data)
  msgpack	Efficient binary serialization format.	import msgpack; data = {'a': 1}; packed_data = msgpack.packb(data)

• Manually convert data to a supported format. Before serialization, convert data to basic JSON types (strings, numbers, booleans, lists, dictionaries, None).

  from datetime import datetime
  
  data = {'timestamp': datetime.now()}
  # Convert datetime to ISO 8601 string before serialization
  data['timestamp'] = data['timestamp'].isoformat()